Fuel management system for mining and fleets
A fuel management system for a mine or a fleet reconciles three things that usually live apart: what was delivered into the tanks, what the tank gauges say is there, and what the pumps dispensed into which vehicle. Mirage connects tanks, pumps, deliveries, transactions and telematics into one system, reconciles them every day per tank and per machine, and flags the variance that is not yet explained. At Transwin, fleet diesel consumption fell 12%.
The problem
Fuel data is fragmented across fixed tanks, mobile tanks, pumps, delivery notes, purchases, vehicles, drivers, GPS, spreadsheets and accounting. A company can usually say how much fuel it bought. Saying where that fuel went is much harder.
Teams struggle to reconcile delivered volumes against what is in the tank, pump transactions against vehicle consumption, and movements across several sites, so discrepancies stay buried in disconnected records until the fuel bill forces the question, weeks after the litres left.
What we deploy
An operational system for managing fuel across one or several sites: fixed and mobile tank tracking, direct connections to pumps and terminal (PTS) equipment, transaction capture, delivery tracking, and reconciliation between theoretical stock, measured stock and actual outflows, down to consumption per truck, vehicle or machine.
It is not a BI dashboard bolted onto a fuel bill and not a simple stock tracker. It is connected to the equipment in the field, combining hardware, a data platform, business rules and an interface built for the teams running the stations, without replacing the systems already in place.
Inputs, outputs, systems
| Reads | Produces | Connected through |
|---|---|---|
| Pump and PTS transactions: volume, pump, time, vehicle or RFID tag, operator | Consumption per vehicle, per machine and per operator; transactions with no vehicle, outside hours or above tank capacity, flagged | Pump controller or PTS export, or a direct connection to the terminal |
| Tank levels from probes or gauges, fixed and mobile tanks | Measured stock per tank, every reading; the variance against theoretical stock, per tank per day | Level probe or gauge interface, or a manual dip entered on site |
| Supplier deliveries: delivery notes, invoices, purchase orders | Delivered volume matched against ordered and received, with the gap flagged before the invoice is paid | ERP or purchasing system, or the delivery note read as a document |
| Engine hours, odometer and GPS from telematics | Litres per 100 km or per engine hour per asset, against the asset's own history and its peers | Telematics provider API or export |
| Asset master, fuel cards, cost centres from the ERP | Fuel cost by site, by fleet and by cost centre, written back for accounting | ERP connector (SAP, Oracle, Microsoft Dynamics, Sage and others) |
| Site rules: tank capacities, authorised vehicles, dispensing hours | The exceptions those rules define, routed to the site or fleet manager with the evidence | Agreed with your team, changed as the site changes |
Outputs are the reconciled record and its exceptions. The method behind the balance, and why a variance is not yet a loss, is in the mine fuel reconciliation guide linked below.
How it works
Mirage connects to pump and PTS data, tank sensors or level probes, refuelling transactions, supplier deliveries, vehicle and GPS or telematics data, purchasing or ERP records, and the RFID or vehicle-ID mappings that tie a transaction to a specific truck or machine. Historical consumption and the rules specific to the site are layered on top.
Every day, per tank, theoretical stock is checked against measured stock and actual outflows; deliveries are matched against what was ordered and received; every transaction is tied back to a vehicle, a pump and a site. Variances, missing volumes, over-consumption and suspicious transactions surface as specific flags with their evidence, instead of disappearing into a spreadsheet.
A daily tank balance, worked through
An illustrative example with invented figures: one 30,000-litre fixed tank, one day. The arithmetic is the one the system runs for every tank, every day.
| Line | Litres | Source |
|---|---|---|
| Opening stock, measured at 06:00 | 18,420 | Level probe |
| Delivery received at 09:40 | +12,000 | Delivery note, matched to purchase order 4471 for 12,000 |
| Dispensed, 47 transactions | −9,860 | Pump controller, every transaction tied to a vehicle |
| Theoretical closing stock | 20,560 | Opening + delivery − dispensed |
| Closing stock, measured at 18:00 | 20,310 | Level probe |
| Variance | −250 | 1.2% of throughput for the day |
| Reading | Within the tolerance set for this tank; tracked, not alerted | A variance of the same sign for five days in a row raises a flag |
A variance is not yet a loss: temperature, probe accuracy, a delivery still settling and a meter drift all move it. The system tracks the sign and the trend per tank and alerts when the pattern, not one reading, says something is wrong.
Operational interface
The operational interface for a system connected to the tanks and pumps themselves, not a fuel dashboard bolted on after the fact.
- Station-by-station overview
- Fixed and mobile tank levels
- Live fuel transactions
- Deliveries and volumes received, matched to orders
- Delivery vs stock vs outflow reconciliation, per tank per day
- Average consumption per vehicle, against its history
- Consumption anomalies and suspicious transactions
- Volume trends over time
- Transaction detail: vehicle, pump, volume, time, site
How do you choose between a fuel card, a pump controller and a fuel management system?
| Fuel cards at public stations | Pump controller on site | Fuel management system | |
|---|---|---|---|
| What it records | Card, litres, station, amount | Vehicle or tag, litres, pump, time | All of that, plus tank levels, deliveries and telematics |
| Reconciles stock | No tank to reconcile | No: it records outflow only | Yes: delivered, in tank and dispensed, per tank per day |
| Consumption per asset | Per card, if cards are not shared | Per vehicle, if tags are not shared | Per asset, against engine hours or kilometres from telematics |
| Catches | Card misuse, after the statement | Transactions outside rules, if configured | Variances, over-consumption, deliveries short of the order, tag misuse, as they happen |
| Fits when | The fleet refuels on the road | One site, one tank, no reconciliation needed | Several tanks or sites, mobile tanks, and a diesel bill that has to be explained |
How it is deployed
Forward-deployed: engineers on your site, with the people who run the stations. The proof of concept usually runs on a month of transaction logs, dips and delivery notes from one site, so you see your own balance before any commitment.
- First reply
- Within two business hours, from an engineer.
- Proof of concept
- On your own data (a telematics export, a transaction log, weighbridge tickets, dispatch records), one week after the first meeting.
- Pilot
- Fixed price, on one site or one fleet, with Mirage engineers deployed with your team. You can stop after it.
- Production
- The first workflow runs in 4 to 6 weeks. A system that spans several sites or a lot of field equipment is scoped at the first meeting and can take longer.
- After go-live
- Mirage operates, monitors and improves the system in production; your team keeps the decision on every exception. More than 200 systems have been connected so far.
- Price
- An annual subscription priced by volume after the pilot, from $10,000 to launch and $1,000 a month, quoted in the first conversation.
A single fixed station with existing PTS and telematics connects fastest; a network of fixed and mobile tanks with mixed hardware is scoped at the first meeting. The method behind the balance, from the per-tank reconciliation to the variance per machine, is set out in the mine fuel reconciliation guide.
FAQ
See what Mirage can run in your operations
Get a custom automation plan
Tell us where your operations lose time. We send back a written plan: which workflows are worth automating, what a deployment looks like on your systems, and the shortest path to production. No commitment.
Prefer to talk it through? Book a call